Google Research

Structured Logging: Crafting Useful Message Content

  • Vladimir Legeza
  • Anton Golubtsov
  • Betsy Beyer
:login;, vol. Summer 2019, Vol. 44, No. 2 (2019)

Abstract

In the context of logging, the word "structured" typically refers to the way log records are represented in a machine-readable format, such as JSON or XML. In this article, we focus on another aspect of logging structure: the message content.

Computing today offers several automated ways of collecting, delivering, and processing log records from different types of systems. But modern technologies are not supportive if the information describing a specific event is insufficient or otherwise not helpful.

To approach this topic, it's useful to understand the most common logging issues, why they occur and possible solutions. By discussing some representative use cases, we aim to provide practical insights and approaches to improving the structure of your logs. As with most advice, our proposed solutions are just one way of approaching a problem space—feel free to either apply our suggestions wholesale or pick and choose the pieces that suit your needs.

Learn more about how we do research

We maintain a portfolio of research projects, providing individuals and teams the freedom to emphasize specific types of work